Leading efforts to build a distributed mutable file system to support data syncing between warehouses like Snowflake and Databricks with amplitude.

This system is like a database processing inserts, updates, and deletes for >2M events / sec.

• Co-led the development and release of a new batch processing pipeline that reduced costs by 15% compared to the realtime pipeline and helped onboard 25 new clients with >50B events monthly data volume. I was overseeing design and implementation, performance improvements, testing, and customer onboarding.

• Collaborated with team member to build an internal monitoring tool that enhanced data collection transparency, tracking 5M+ metrics / minute. Took on additional frontend development responsibilities to ensure timely delivery of this tool which is used by 30+ CSMs for debugging customer issues.

• Led the design and implementation of a new batch-based billing system for tracking monthly user count to support a new billing plan.

• Launched social media platform integrations enabling clients to search across 5M influencers globally.

• Developed library for managing task dependencies that enhanced durability and throughput for social media micro services by 20%.

• Designed scalable data pipelines, and cloud infrastructure to process >10M events daily.

• Led the core Data Science team of 3 engineers to build proprietary Computer Vision and NLP models that power the influencer search engine.

• Trained and grew the engineering team from 2 to 20 across 3 years.

• Optimized cloud spend across AWS, GCP, Azure, IBM thereby saving the company $100K.

• Architected a service that maps social media profiles across different platforms using face, location, and biography analysis, thereby mapping 200k profiles with 96% accuracy.

• Developed a magic-link authentication service to enable easy onboarding of influencers.

• Developed a real-time Influencer recommender engine that improved usage and retention by 20%.

• Started an EdTech company that helps users learn their choice of language for day-to-day micro-tasks.

• Successfully raised pre-seed round from Entrepreneur First; developed prototype for 3 clients, and validated the idea with over 100 users and language schools.